Jeff Clune

@jeffclune

Research: AI, ML, Deep RL, Deep Learning. Senior Research Manager, Uber AI Labs. Harris Associate Professor, University of Wyoming.

Vrijeme pridruživanja: srpanj 2010.

Tweetovi

Blokirali ste korisnika/cu @jeffclune

Jeste li sigurni da želite vidjeti te tweetove? Time nećete deblokirati korisnika/cu @jeffclune

  1. Prikvačeni tweet
    8. sij 2019.

    Introducing POET: it generates its own increasingly complex, diverse training environments & solves them. It automatically creates a learning curricula & training data, & potentially innovates endlessly! By w/ &

    Prikaži ovu nit
    Poništi
  2. 31. sij

    Awesome to watch you become so successful Anh! And always a pleasure to see you. Congratulations!!

    Poništi
  3. 30. sij

    Thank you very much Nando! I am glad to hear you like it.

    Poništi
  4. 30. sij

    Excited to give a talk today on ANML: Learning to Continually Learn. Recording available soon. If you are curious about ANML now, here is a NeurIPS talk I gave that covers it (amongst other things). Talk today is at the SF summit

    Poništi
  5. 29. sij

    This makes me happy. Great black/whiteboarding!

    Poništi
  6. 28. sij

    Just to be clear, I was not involved in the great team that did this work. Congratulations to that team!

    Prikaži ovu nit
    Poništi
  7. 28. sij
    Prikaži ovu nit
    Poništi
  8. 27. sij

    An enduring consequence of having watched MadMen years ago is that every time I take notice of any bit of marketing, I immediately see Don Draper making the pitch for that idea in a 60s suit and conference room.

    Poništi
  9. proslijedio/la je Tweet
    24. sij
    Poništi
  10. 21. sij

    Please consider contributing to this exciting workshop we are organizing!

    Poništi
  11. 17. sij

    New bounding boxes image dataset on 10,000 camera trap images from 23 classes from Eastern North America. The most common classes are “American Crow”, “American Black Bear”, and “Dog”. Thanks Yousif H, Kays R, Zhihai H & Dan Morris

    Poništi
  12. 16. sij

    I could not agree more. What a colossal waste of the extremely valuable time of talented people. This is a self-inflicted wound caused by shortsighted bureaucrats.

    Poništi
  13. 14. sij

    Excited to give a talk GTC. Title: AI-GAs: AI-Generating Algorithms, an Alternate Paradigm for Producing General Artificial Intelligence. March 22-26, 2020. AI-GA quick summary , paper: . Affiliations out of date.

    Poništi
  14. 14. sij

    This work is indeed very cool! It was inspired in part by our work on Evolving Soft Robots (video: ). Review of the field and work since by , 1st author on the new amazing PNAS work (w & ):

    Poništi
  15. proslijedio/la je Tweet
    13. sij

    We are excited to announce our workshop 'Beyond "Tabula Rasa" in RL (BeTR-RL): Agents that remember, adapt, and generalize' w/ Ignasi Clavera, Kate Rakelly, , , CfP & more infos:

    Prikaži ovu nit
    Poništi
  16. proslijedio/la je Tweet
    13. sij
    Odgovor korisnicima i sljedećem broju korisnika:

    There's value in a generative approach and there's also value in looking for first principles. One informs the other. This combination is what I find to be promising. AGIs will be grown, but they will be grown on the basis of first principles.

    Poništi
  17. proslijedio/la je Tweet
    12. sij
    Odgovor korisnicima

    This is a chart of the best-studied, most understandable metabolic pathways. These are considered one of the simplest and most tractable things to study in molecular biology. Now imagine how hideously complex and intractable the hard questions are, and then multiply it by 10.

    Poništi
  18. 9. sij

    They both also discussed how finding the right minimal set of priors to enable learning powerful AI is key. I agree. Their answer [At the 2:58:48 mark.] Thanks to for inviting me to ask a question and organizing a great debate! 4/4

    Prikaži ovu nit
    Poništi
  19. 9. sij

    ' take: I pretty much agree with Yoshua. In principle we know that evolution is a mechanism that is powerful enough to evolve minds because it evolved our minds; having the machine do the work that stands in for evolution would be great...totally worth exploring” 3/4

    Prikaži ovu nit
    Poništi
  20. 9. sij

    Yoshua: "Science is an exploration, we don’t know what is going to work, [manual & AI-GAs] are two different directions and they can co-exist in a harmonious way”. (2/4)

    Prikaži ovu nit
    Poništi
  21. 9. sij

    Yoshua Bengio's take on AI-Generating Algorithms in the : AI-GAs are "an ambitious research agenda...very interesting and important investigation, very exciting." More from Yoshua and Gary Marcus below (1/4)

    Prikaži ovu nit
    Poništi

Čini se da učitavanje traje već neko vrijeme.

Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.

    Možda bi vam se svidjelo i ovo:

    ·